Online reservation and allocation management for integrated parking and charging lots

Abstract

<p>The global electric vehicle (EV) ownership exceeds 26 million in 2022, marking a 60% increase from 2021 and more than five times that of 2018. To meet the ensuing surge in charging demand, many local governments have exerted tremendous efforts to bolster the construction of charging facilities. Despite the growing supply, EV drivers still struggle to find available charging spaces, as they are often occupied by non-charging vehicles or EVs that have finished charging. This phenomenon undermines the interests of charging facilities operators and user satisfaction, hindering the promotion of electric vehicles. While a complete prohibition on parking at charging spaces can circumvent parking at charging spaces, it may lead to underutilization of charging spaces during periods of low charging demand. This study aims to tackle this problem through a parking reservation system embedded with admission and allocation controls. Parking reservation technology has been proven effective in reducing cruising and congestion. It leverages the Internet of Things (IoT) to collect real-time information on parking space availability. Users can access this information and reserve parking or charging spaces in advance, mitigating the risk of arriving at a parking lot only to find no available spaces. Upon receiving a reservation, the operator determines whether to accept or reject it. Despite the extensive research on traditional parking systems, there is a lack of a systematic framework for modeling and optimizing the planning and operation of IPCLs. Managing IPCLs presents intricate challenges stemming from both demand and supply aspects. On the demand side, users have distinct service requirements in terms of space types, duration, and desired charging amount (for charging users), and their arrival patterns vary throughout the day. On the supply side, charging spaces can serve both parking and charging needs, while parking spaces only accommodate parking demand. The heterogeneous, uncertain, and time-varying nature of demand, along with the asymmetric substitutability of supply, necessitates the development of flexible and adaptive strategies to efficiently allocate parking and charging spaces. In other words, the operator must determine whether, when, and to what extent charging spaces can be allocated for parking purposes to maximize the overall revenue of the parking lot. This study considers an IPCL with partial parking spaces equipped with charging facilities. The IPCL adopts a reservation system where users can reserve a space by indicating the space type, start time, end time, and current and desired states of charge (for charging requests only). The operator evaluates the request based on the availability of spaces at the requested time and decides whether to accept it. If accepted, the operator earns a profit and allocates a compatible space. We employ a Markov decision process (MDP) model to capture this sequential decision-making process. Given the high complexity of deriving the optimal strategy, we introduce two decomposition-based strategies that yield fast admission and allocation decisions based on real-time space availability and demand forecasts. Numerical experiments demonstrate the effectiveness of the strategies across various demand scenarios, with revenue increases of 18.8% and 11.3% compared to the cases with no intervention and full permission for parking at charging spaces, respectively. Additionally, the selection of charging powers and the number of charging spaces in the parking lot are optimized. The findings indicate that different charging powers are suitable for different demand scenarios, and fast charging tend to be ideal for high charging demand. These observations provide guidance for the planning and operation of integrated parking and charging lots.</p>